IDF21-0470 Neither plasma glucose at admission nor in-hospital hyperglycaemia predict poor clinical outcomes in COVID-19 patients

Abstract

Background: Diabetes mellitus and hyperglycaemia are associated with poor outcomes in COVID-19, however, the predictors of clinical outcomes are variable across all the available studies. Besides, data from the Indian subcontinent are limited in this regard. Aim: To compare clinical outcomes in COVID-19 patients with and without hyperglycaemia and delineate the clinical and biochemical predictors in a cohort of adults hospitalized at a tertiary care institution in India. Method: In this prospective single-center cohort study, consecutive adults (≥18 years) admitted with PCR-proven diagnosis of COVID-19 between December 2020 and March 2021 were recruited. Patients who had received glucocorticoids prior to present admission were excluded. Complete blood count, biochemical/coagulation panel, random plasma glucose (RPG), C-reactive protein (CRP), interleukin-6, ferritin, lactate dehydrogenase (LDH), creatine kinase, troponin I, pro-brain natriuretic peptide, procalcitonin, glycated hemoglobin (HbA1c), thyroid function test, cortisol and 25-hydroxyvitamin D were estimated within 24 h of admission. Patients with a prior history of type 2 diabetes (T2D) were categorized as pre-existing diabetes (preDM). In those without a prior history of diabetes, an admission HbA1c ≥6.5% was categorized as new-onset diabetes (newDM); a normal HbA1c with admission RPG > 140 mg/dl was classified as new-onset hyperglycaemia (newGly). Patients with normal HbA1c/RPG at admission were grouped as having normoglycaemia (normoGly). In-hospital hyperglycaemia was defined as any 2 consecutive random capillary blood glucose (CBG) values >180 mg/dl. The primary outcome was a composite of in-hospital death and/or need for invasive mechanical ventilation (death/IMV). Results: A total of 440 COVID-19 patients were included (mean age 51.1 ± 16.9 years; M:F=1.5:1). At admission, 36%, 13%, 13% and 38% of patients had preDM, newDM, newGly and normoGly, respectively. Overall, 62% had hyperglycaemia at admission (preDM+newDM+newGly). One hundred patients (22.7%) meet with death/IMV. Patients with hyperglycaemia at admission had a higher chance of death/IMV than those with normoGly (78% vs. 28%, p = 0.017). No significant difference in death/IMV was observed in those with and without in-hospital hyperglycaemia (62% vs. 38%, p = 0.063). On binomial logistic regression, the following clinical/biochemical parameters at admission were predictive of death/IMV: SpO2 (OR 0.92, 95% CI: 0.85–1.00, p = 0.050), requirement of non-rebreather mask as compared to non-requirement of oxygen (OR 4.59, 95% CI: 1.67-12.64, p = 0.003), total leukocyte count (OR 1.10, 95% CI: 1.02–1.19, p = 0.013), platelet count (OR 0.995, 95% CI: 0.991–0.998, p = 0.003), albumin (OR 0.30, 95% CI: 0.13-0.69, p = 0.005), thyroxine (T4) (OR 0.82, 95% CI: 0.68–0.99, p = 0.045) and cortisol (OR 1.002, 95% CI: 1.000–1.003, p = 0.008).Notably, RPG at admission, in-hospital hyperglycaemia, age, sex, use of steroids/remdesivir/tocilizumab in hospital and choice of in-hospital anti-diabetic drugs were not predictive of death/IMV on binomial logistic regression analyses. Discussion: Although COVID-19 patients with hyperglycaemia at admission tend to have poor clinical outcomes, neither plasma glucose at admission nor in-hospital glycemic status are reliable prognostic markers. Factors other than ambient plasma glucose might be operational in dictating poor clinical outcomes in COVID-19 patients with hyperglycaemia and diabetes mellitus.